Most individuals with acquired language impairments experience difficulty in both written and spoken language. However, research on the rehabilitation of written language has lagged behind that designed to improve spoken language. A number of studies document the therapeutic value of behavioral treatment for acquired impairments of spelling (agraphia), but the evidence is not adequate to guide treatment selection for a particular patient or to inform expectations regarding the magnitude of change effected by treatment. Recent advances in the understanding of the cognitive processes and neural substrates that support written language provide a framework to guide treatment approaches for agraphia. Our work to date has shown that spelling involves lexical-semantic and phonological processes engaging relatively distinct neural substrates that can be differentially impaired by focal brain damage. With regard to the treatment of acquired impairments of spelling, we hypothesize the following: (1) Individuals with damage to left perisylvian language areas will have impaired phonological abilities with relative preservation of orthographic knowledge, and their spelling performance will improve in response to phonological treatment. (2) Individuals with damage to left posterior extrasylvian language areas will have impaired orthographic knowledge with relatively spared phonological abilities, and their spelling will improve in response to interactive treatment to promote optimal use of residual lexical-semantic and phonological knowledge. (3) Individuals with extensive left hemisphere damage will have significant impairment to both phonological and lexical-semantic processes, and their spelling will improve in response to a lexical treatment approach that involves releaming of specific orthographic representations. To test these hypotheses, we will examine three treatment protocols for acquired agraphia that specifically target lexical, phonological, or interactive spelling procedures. Single-subject experiments will be replicated across participants, with meta-analyses implemented for each treatment. We anticipate that this research will provide evidence-based treatment protocols for acquired impairments of spelling and guidelines for optimal candidacy and implementation. The findings should further clarify understanding of the cognitive processes and neural substrates that support spelling and recovery of function following brain damage.